SwRI 2003, Dec 16 Triton Atmosphere
Triton's atmosphere: energy crisis
Leslie Young (SwRI)Glenn Stark (Wellesley)Ron Vervack (JHU/APL)
SwRI 2003, Dec 16 Triton Atmosphere
Triton atmosphere overview
2706 km diameter
SwRI 2003, Dec 16 Triton Atmosphere
Triton atmosphere overview
2706 km diameter
Tropospheric hazes, plumes, w
inds
SwRI 2003, Dec 16 Triton Atmosphere
Triton atmosphere overview
2706 km diameter
Nitrogen frost transport
Tropospheric hazes, plumes, w
inds
SwRI 2003, Dec 16 Triton Atmosphere
Triton atmosphere overview
2706 km diameter
Nitrogen frost transport
Tropospheric hazes, plumes, w
inds
UVS occultation (<
80 nm)
SwRI 2003, Dec 16 Triton Atmosphere
Triton atmosphere overview
2706 km diameter
Nitrogen frost transport
Tropospheric hazes, plumes, w
inds
UVS occultation (>
80 nm)
UVS occultation (<
80 nm)
SwRI 2003, Dec 16 Triton Atmosphere
Triton atmosphere overview
2706 km diameter
Nitrogen frost transport
Tropospheric hazes, plumes, w
inds
UVS occultation (>
80 nm)
UVS occultation (<
80 nm)
Radio occultation
SwRI 2003, Dec 16 Triton Atmosphere
UVS occultation at a glance
600Wavelength (Å)
700 800 900 10000
200
400
600
800
Alti
tude
(km
)
0.0
0.2
0.4
0.6
0.8
1.0
Tra
nsm
issi
on
N2 electronic bands
N2 ionization Other
SwRI 2003, Dec 16 Triton Atmosphere
Previous Voyager-based models
0
20
40
60
80
100
120
0 100 200 300 400 500 600 700 800
to lower atmosphereFlux of thermospheric energy
Altitude (km)
Thermospheric heating peaks at 350-400 km•Solar EUV and UV heating•Precipitation of electrons from Neptune's magnetosphere•Chemical heating
36
38
40
42
44
0 10 20 30 40 50
Altitude (km)
to lower atmosphereFlux of thermospheric energy
•Turbulent heating & cooling•Negligible radiation and condensation
Tropospheric cooling peaks at 8-15 km
SwRI 2003, Dec 16 Triton Atmosphere
Everyone clamors for new N2 line data
“The N2 densities can be measured to much lower altitudes by the use of the electronic bands in the 1000Å region.” -Broadfoot et al 1989
“…laboratory measurements of N2 ultraviolet absorption properties are urgently needed to permit analysis of the UVS occultation data in the 800-1000Å region, which senses the atmosphere down to ~50 km.” - Yelle et al. 1995
“The temperature profile in the 310 km gap (10-320 km) is contained in the unanalyzed N2 band absorption signatures in the UVS solar occultation data between 850 and 1000Å. Until the necessary laboratory data is available to perform this analysis, we must rely on theoretical models to infer the temperature profile connecting the tropopause to 450 km.” - Strobel et al. 1995
“Below 450 km occultation data is available in the 660-1000Å region where N2 absorbs strongly in many discrete electronic bands... We are not able to accurately model the absorption of sunlight to derive the structure of Triton’s atmosphere from the surface to 450 km because of inadequate molecular cross section data.” - Stevens et al. 1992
SwRI 2003, Dec 16 Triton Atmosphere
Limitations of old N2 line data • Line positions well known• Measured f-values (e.g. strengths) for bands, so
used Hönl-London factors to derive individual line strengths. But this region is full of perturbations, so the Hönl-London factors misestimate lines strengths by factors of 2-3. – Hönl-London factors are the ratios between the strengths for
individual lines and the band. For a given band, these factors are simple functions of the rotational states.
• Predissociation rates (and therefore widths) are largest uncertainty. (Lines are described by Voigt profiles, with Gaussian width from temperature, Lorentz width from predissociation)
SwRI 2003, Dec 16 Triton Atmosphere
New N2 line data • Collaborator Glenn Stark (Wellesley College)• Surveyed all bands, 80-100 nm, with multiple
scans, multiple pressures, room temperatures (so colder Triton temperatures will not introduce hot overtones).
• 17 bands available for this analysis (up from 4)– N2 photoabsorption measured by Stark and others at the Photon
Factory synchrotron radiation facility at the High Energy Accelerator Research Organization in Tsukuba, Japan, 1997, 1998, 1999
– Predissociation widths from Stark’s measurements, or from Ubachs (e.g., Ubachs, W. 1997. Predissociative decay of the c4’ 1∑u
+ v=0 state of N2, Chem. Phys. Lett. 268, 201.)– In most cases, measured individual line strengths and
predissociation widths. (new this year)– Use cross sections for T=95 K
SwRI 2003, Dec 16 Triton Atmosphere
N2 electronic bands used (new) Mean wavelength Total strength Dissociation Width band (Å) (Å cm2) (mÅ) ---- --------------- -------------- ------------------ b(8) 935.3 2.8E-18 0.500 b’(4) 938.0 1.4E-17 1.600 c3(1) 938.8 2.8E-16 0.577
b(7) 942.6 1.3E-16 0.150 b’(3) 944.8 5.6E-19 0.000 o(0) 946.3 2.0E-18 0.200 b(6) 949.4 3.1E-17 0.140 b’(2) 951.2 2.8E-19 0.000 b(5) 955.3 2.8E-17 0.240 b’(1) 958.3 1.6E-17 0.070 c4’(0) 958.6 1.1E-15 0.073
c3(0) 960.4 4.0E-16 0.670
b(4) 965.9 5.6E-16 2.900 b(3) 972.3 3.7E-16 38.727 b(2) 979.1 1.7E-16 5.165 b(1) 985.8 6.9E-17 0.050 triple 989.6 7.6E-21 5000.000 b(0) 992.0 1.9E-17 1.500
SwRI 2003, Dec 16 Triton Atmosphere
Voyager (1989) occultation data
0
0.2
0.4
0.6
0.8
1
0
0.4
0.8
1.2
1.6
2
0 100 200 300 400 500 600 700 800 altitude (km)
UVS Channel 9(576.48 to 585.74 Å)
UVS Channel 49(945.63 to 954.89 Å)
Radio egress
0
0.2
0.4
0.6
0.8
1
0
0.4
0.8
1.2
1.6
2
0 20 40 60 80 100 altitude (km)
Radio egressUVS Channel 49(945.63 to 954.89 Å)
UVS Channel 9(576.48 to 585.74 Å)
SwRI 2003, Dec 16 Triton Atmosphere
Voyager data details• UVS transmission vs. altitude and wavelength
– Use Ron Vervack’s most recent reduction (~1992)– Use only odd channels– Account for wavelength shifts with altitude by shifting the
wavelengths of channels (new; had shifted the spectra, not the wavelengths of the bins)
– Estimate errors from photon count, following Yelle et al 1993 (new; had estimated errors by comparing ingress and egress).
• Radio phase delay vs. altitude– Using most recent reduction: Gurrola, E. M. 1995. Interpretation
of Radar Data from the Icy Galilean Satellites and Triton. Ph.D. Thesis, Stanford University, Fig 6.2, as preserved in Planetary Data System
– Gurrola estimates errors at 0.1 radian; however, errors are highly correlated
– Use the background estimated by Gurrola, taking into account this correlation; this is the background in Gurrola Table 8.3.
SwRI 2003, Dec 16 Triton Atmosphere
N2 cross sections
10-21
10-20
10-19
10-18
10-17
10-16
10-15
10-14
940 945 950 955 960
Cross Section (cm
2)
wavelength (A)
Channel 49
SwRI 2003, Dec 16 Triton Atmosphere
Line-of-sight density details• Nlos from N2 continuum transmission
– Simple Beer’s law
– , where
• Nlos from N2 line transmission– Explicit averaging over wavelength– Consistent with Nlos from N2 continuum transmission
–
• Nlos from radio phase
–
transmission eNlos
F d 0
0
F d0
0
transmission e Nlos
d 0
0
phase delay 4STPNL
Nlos
SwRI 2003, Dec 16 Triton Atmosphere
New line-of-sight number density
1015
1016
1017
1018
1019
1020
1021
1022
1023
0 100 200 300 400 500 600 700 800altitude (km)
Radio egress
UVS Channel 49
UVS Channel 9
1021
1022
1023
0 20 40 60 80 100altitude (km)
Radio egress
SwRI 2003, Dec 16 Triton Atmosphere
What can we do?• Weight transmission by solar flux
– Want a good solar spectrum with F10.7=178 W/m2/Hz
• Smooth spectra to match instrumental resolution– 13 Å triangular smoothing function
• Use multiple channels• Use a temperature-dependent cross section
– but the cross sections at 500 km should be at 95 K
• For now, look at last year’s self-consistent reduction...
SwRI 2003, Dec 16 Triton Atmosphere
10 15
10 16
10 17
10 18
10 19
10 20
10 21
10 22
10 23
0 100 200 300 400 500 600 700 800
altitude (km)
Radio egress
UVS Channel 49(957.90 to 967.16 Å)
UVS Channel 9(587.53 to 596.79 Å)
10 21
10 22
10 23
0 20 40 60 80 100
altitude (km)
Radio egress
Old line-of-sight number density
SwRI 2003, Dec 16 Triton Atmosphere
Nlos interpretation details• Fit UVS and radio with (small-planet) isothermal
models
–
• Upper atmosphere– All of the SNR>5 UVS Nlos can be fit with single temperature– T=104 K (c.f. Krasnopolsky et al. 1993; T=102±3 K)
• Lower atmosphere– All of the SNR>5 radio Nlos consistent with single temperature– T=44 K (c.f. Gurolla 1995; T=42±8 K)
• Comparison with 1997– Calculate Nlos from temperature profile (Elliot et al. 2000)– Nlos has increased between 1989 and 1997
Nlos =Nlos(ro)rr0
⎛
⎝ ⎜
⎞
⎠ ⎟
3/ 2
exprH
−r0H0
⎡
⎣ ⎢
⎤
⎦ ⎥ 1+
98
Hr
−H0r0
⎛
⎝ ⎜
⎞
⎠ ⎟
⎡
⎣ ⎢ ⎢
⎤
⎦ ⎥ ⎥
SwRI 2003, Dec 16 Triton Atmosphere
Nlos interpretation
10 15
10 16
10 17
10 18
10 19
10 20
10 21
10 22
10 23
0 100 200 300 400 500 600 700 800
1989 UVS N2 lines
1989isothermalloweratmosphere(43 K) 1989 isothermal
upper atmosphere (104 K)
altitude (km)
1989 Radio
1989 UVS N2 continuum
1997 Stellar Occultation
10 15
10 16
10 17
10 18
10 19
10 20
10 21
10 22
10 23
0 50 100 150 200
1989 UVS N2 lines1989
isothermalloweratmosphere(43 K) 1989 isothermal
upper atmosphere (104 K)
altitude (km)
1989 Radio1997 Stellar Occultation
SwRI 2003, Dec 16 Triton Atmosphere
Temperature profile details• Invert the “spliced isothermal” Nlos profile from
above– Use the “small planet” Abel transform
–
• Lower atmosphere (<90 km): colder in 1989 than in 1997– As reported by Elliot et al. 2000– Higher conductive flux will affect thermal models of the 1997
occultation, which are currently unsatisfactory.
n(r) =−1
πr2dsNlos(s)
dssds
s2 −r20
∞∫
SwRI 2003, Dec 16 Triton Atmosphere
Temperature profiles
0
20
40
60
80
100
120
0 100 200 300 400 500 600 700 800
Altitude (km)
1997stellar occ
1989 UVS and radio occs, new temperature profile (example)
1989 UVS and radio occs, previous temperature profiles
40
50
60
70
80
90
100
110
0 50 100 150 200
Altitude (km)
1997stellar occ
1989 occs, new profile
1989 occs, previous profiles
SwRI 2003, Dec 16 Triton Atmosphere
What’s the energy source near 100 km?
• Implied heating differs from previous models– Stevens et al. 1992; Yelle et al. 1991
• Main heating is below 150 km – c.f. Stevens et al 1992; 300-400 km
• Gradients may be as high as 1.4 K/km for fluxes as high as 8x10-3 erg/cm2/s– c.f. Broadfoot et al. 1989, Yelle et al. 1991; 1x10-3 erg/cm2/s
• Heating by energetic particle precipitation?– That can reach lower altitudes, but the total energy is a problem
• Thermal models may not split the atmosphere into “upper” and “lower”– Page charges will be larger, author lists will be longer
SwRI 2003, Dec 16 Triton Atmosphere
An endemic problem? Triton’s 1989 atmosphere is warmer than models, as
is:• Pluto’s 1989 atmosphere (Lellouch 1994; Strobel et al. 1996)
• Triton’s 1997 atmosphere (Elliot et al. 2000)
• Jupiter, Uranus near 1 µbar
SwRI 2003, Dec 16 Triton Atmosphere
Conclusions• We are close to realizing the promise of the N2
electronic bands (but more work still to be done)• Supporting evidence for changes in Triton’s
atmosphere between 1989 and 1997• So far, evidence for a major puzzle in the energetics
of Triton’s atmosphere near 100 km altitude (0.25 µbar)
• The solution to this puzzle may have larger implications in the outer solar system.